-
Notifications
You must be signed in to change notification settings - Fork 10
Description
Hi; during the build process through the example for changenet, the dockerfile build fails after the build command:
cd builder/samples
docker compose up --build ms-changenetAnd running it gives this:
[+] Building 431.2s (8/13)
=> [internal] load local bake definitions 0.0s
=> => reading from stdin 647B 0.0s
=> [internal] load build definition from Dockerfile 0.0s
=> => transferring dockerfile: 2.74kB 0.0s
=> [internal] load metadata for nvcr.io/nvidia/deepstream:8.0-triton-multiarch 0.0s
=> [internal] load .dockerignore 0.0s
=> => transferring context: 2B 0.0s
=> CACHED [builder_base 1/8] FROM nvcr.io/nvidia/deepstream:8.0-triton-multiarch 0.0s
=> [internal] load build context 0.0s
=> => transferring context: 25.60kB 0.0s
=> [builder_base 2/8] RUN --mount=type=cache,target=/root/.cache/pip pip install requests==2.32.3 python-multipart==0.0.9 aiofiles==23.2.1 a 422.5s
=> ERROR [builder_base 3/8] RUN --mount=type=cache,target=/root/.cache/pip pip install inferencemodeltoolkit==1.0.11 --extra-index-url https://urm.nvidia.com/ 8.5s
------
> [builder_base 3/8] RUN --mount=type=cache,target=/root/.cache/pip pip install inferencemodeltoolkit==1.0.11 --extra-index-url https://urm.nvidia.com/artifactory/api/pypi/nv-shared-pypi/simple:
0.675 Looking in indexes: https://pypi.org/simple, https://urm.nvidia.com/artifactory/api/pypi/nv-shared-pypi/simple
0.870 WARNING: Retrying (Retry(total=4, connect=None, read=None, redirect=None, status=None)) after connection broken by 'NewConnectionError('<pip._vendor.urllib3.connection.HTTPSConnection object at 0x7d0f710a2cc0>: Failed to establish a new connection: [Errno -5] No address associated with hostname')': /artifactory/api/pypi/nv-shared-pypi/simple/inferencemodeltoolkit/
1.375 WARNING: Retrying (Retry(total=3, connect=None, read=None, redirect=None, status=None)) after connection broken by 'NewConnectionError('<pip._vendor.urllib3.connection.HTTPSConnection object at 0x7d0f710a2bd0>: Failed to establish a new connection: [Errno -5] No address associated with hostname')': /artifactory/api/pypi/nv-shared-pypi/simple/inferencemodeltoolkit/
2.379 WARNING: Retrying (Retry(total=2, connect=None, read=None, redirect=None, status=None)) after connection broken by 'NewConnectionError('<pip._vendor.urllib3.connection.HTTPSConnection object at 0x7d0f710a2de0>: Failed to establish a new connection: [Errno -5] No address associated with hostname')': /artifactory/api/pypi/nv-shared-pypi/simple/inferencemodeltoolkit/
4.384 WARNING: Retrying (Retry(total=1, connect=None, read=None, redirect=None, status=None)) after connection broken by 'NewConnectionError('<pip._vendor.urllib3.connection.HTTPSConnection object at 0x7d0f710a2fc0>: Failed to establish a new connection: [Errno -5] No address associated with hostname')': /artifactory/api/pypi/nv-shared-pypi/simple/inferencemodeltoolkit/
8.390 WARNING: Retrying (Retry(total=0, connect=None, read=None, redirect=None, status=None)) after connection broken by 'NewConnectionError('<pip._vendor.urllib3.connection.HTTPSConnection object at 0x7d0f710a31d0>: Failed to establish a new connection: [Errno -5] No address associated with hostname')': /artifactory/api/pypi/nv-shared-pypi/simple/inferencemodeltoolkit/
8.422 ERROR: Could not find a version that satisfies the requirement inferencemodeltoolkit==1.0.11 (from versions: none)
8.424 ERROR: No matching distribution found for inferencemodeltoolkit==1.0.11
------
Dockerfile:36
--------------------
35 |
36 | >>> RUN --mount=type=cache,target=/root/.cache/pip \
37 | >>> pip install inferencemodeltoolkit==1.0.11 --extra-index-url https://urm.nvidia.com/artifactory/api/pypi/nv-shared-pypi/simple
38 |
--------------------
failed to solve: process "/bin/sh -c pip install inferencemodeltoolkit==1.0.11 --extra-index-url https://urm.nvidia.com/artifactory/api/pypi/nv-shared-pypi/simple" did not complete successfully: exit code: 1
I think this may have something to do with using the urm.nvidia source for the package url since from what I've seen online it seems that this is an internal resource and isn't meant to be accessible to the public. I saw that this exists in seemingly three places in the project:
The changenet dockerfile at:
inference_builder/builder/samples/changenet/Dockerfile
And the vila dockerfile at:
inference_builder/builder/samples/vila/Dockerfile
as well as in the root level .Dockerfile.builder file (which has its own sort of confusing stuff like including a conflicting python version from the overall README.md)
EDIT:
Additionally, there is a dependency on this "inferencemodeltoolkit" at:
inference_builder/templates/api_server/triton/responder.jinja.py
which manifests in the location (inside of the unzipped changenet.tgz):
/server/responder.py